|
Items in cart [4] |
Questions? Call 888-624-8373 |
| Copyright © 2009. National Academy of Sciences. All rights reserved. Terms of Use and Privacy Statement |
Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter.
Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.
OCR for page 173
Issues in Risk Assessment
Appendix G
Informal Search for ''Supercarcinogens"
As discussed in the text of the report, the graph of TD50 versus MDT can be divided empirically into three regions (Figure 1). By observation, most carcinogens are in a narrow band (Region B); few are above (Region A) or below (Region C) this band. The absence of carcinogens from Region A might be partly or entirely artifactual. Any carcinogen whose biologic properties would place it in Region A would have such low potency (such a high TD50) that it would give a very weak response even if tested at the MTD; hence, it would not be recognized as carcinogenic and would not be included in the CPDB. Thus, no conclusions can be drawn from the apparent emptiness of Region A.
The absence of carcinogens from Region C, however, is not obviously artifactual. The argument put forward by Rieth and Starr (1989b) that any carcinogens that belong properly in Region C would yield 100% tumor incidence in a conventional bioassay, so a finite TD50 could not be calculated, is incorrect. For some chemicals that yield 100% tumor incidence, the CPDB includes a 99% upper confidence limit on the TD50 (Gold et al., 1986a); for others, tumor incidence less than 100% can be observed in bioassays conducted at lower doses or for periods shorter than a lifetime. Thus, if some chemicals truly belong in Region C, they should be detectable, and it should be possible to derive numerical estimates of potency for at least some of them.
Nevertheless, the committee identified several types of bias that might
OCR for page 174
Issues in Risk Assessment
FIGURE 1
OCR for page 175
Issues in Risk Assessment
hypothetically have led to the exclusion of carcinogens that properly belong in Region C from the CPDB and from studies based on the CPDB. For purposes of discussion, we refer to these hypothetically excluded carcinogens that properly belong in Region C as "supercarcinogens"—defined for this purpose as carcinogens whose TD50s lie below the 95% error bound on the regression line in Figure 1, i.e., less than about MTD/7. One hypothetical source of bias is that some of these agents were identified as potent carcinogens long ago and, being well known as such, were never tested in up-to-date bioassays and so did not have met the inclusion criteria of the CPDB. Another possible source of bias is that some of the agents, if tested at the MTD, yielded tumors in very short periods and were excluded from the CPDB because of early termination of the studies. Krewski's criteria for selecting chemicals from the CPDB for analysis could have introduced other, more subtle biases.
To investigate whether those hypothetical biases are important, the committee conducted a search for supercarcinogens that exist but have been excluded from the CPDB or from Krewski's analysis. The search was necessarily informal, because there is no systematic compilation of carcinogenic potencies other than the CPDB. The committee's approach was to compile a list of candidate chemicals with various criteria and then to review the data on them to explore whether they might fall into Region C, either according to the inclusion criteria and calculation procedures of the CPDB and of Krewski or according to modified criteria and procedures. The results of the search are reported in this appendix.
CRITERIA AND CANDIDATE CHEMICALS
The following criteria were used to identify candidate chemicals for this study:
"Classical" carcinogens—identified before 1965 and not subjected to modern bioassays.
Agents that induced tumors in less than 6-months and might never have been tested in a lifetime bioassay.
Other agents that are generally recognized as "potent" carcinogens and might never have been formally tested for carcinogenicity.
OCR for page 176
Issues in Risk Assessment
Agents that have been tested over an unusually wide range of doses and are believed to be effective at doses below the MTD by a factor of at least 100.
Other agents nominated by committee members.
On the basis of those criteria, the committee selected 18 candidate agents for study (Table G-1).
TABLE G-1 Chemicals and Other Agents Considered in this Study
Agent
Criteria for Inclusion in the Studya
2-Acetylaminofluorene
D
Acrylonitrile
E
Benzidine
B (parent compound of benzidine dyes)
Benzo[a]pyrene
A
1,3-Butadiene
D
Carbon tetrachloride
C
C.I. Direct Black 38
B
C.I. Direct Blue 6
B
C.I. Direct Brown 95
B
Dibenz[a,h]anthracene
A
Dimethyl sulfate
C
Ethylene dibromide
B
Ethylene oxide
E
Ethylnitrosourea
C
Methyl bromide
B
MOCA
E
Plutonium
A,B,C,D, (most potent member of class of radionuclides)
Vinyl Chloride
D
aSee text.
OCR for page 177
Issues in Risk Assessment
DATA
Twelve of the candidates were already included in the CPDB and in Krewski's study (upper portion of Table G-2). For two other agents (benzo[a]pyrene and 1, 3-butadiene), the committee identified dose--response data that could be analyzed quantitatively (Tables G-3 and G-4). For another agent (vinyl chloride), the committee identified an ingestion study that gave results (Table G-5) markedly different from those of the inhalation study included in the first part of Table G-2. The ingestion study (Feron et al., 1981) appears to have met the inclusion criteria of the CPDB, and it is not clear why it was not included in the CPDB. The data in Tables G-3, G-4, and G-5 were analyzed by Krewski with the same methods as those used in his workshop paper, and the resulting estimates of TD50 are tabulated in the lower portion of Table G-2.
For four agents listed in Table G-1, comparable numerical estimates of carcinogenic potency could not be obtained, for the following reasons:
Dimethyl sulfate. The only reported studies are unsuitable for quantitative analysis, but show tumors at the MDT and MDT/2 (IARC, 1974).
Dibenz[a,h]anthracene. The only reported studies are unsuitable for quantitative analysis (ATSDR, 1990).
Methyl bromide. Data purporting to show induction of forestomach tumors within 90 days (Danse et al., 1984) have been discredited (EPA, 1986; Reuzel et al., 1991).
Plutonium. Dose data on this and other radionuclides are not commensurable with those customarily applied to chemical carcinogens. For plutonium, the radiation dose that causes early death (within 1.5 years) due to radiation pneumonitis and pulmonary fibrosis in animals exposed by inhalation is about 45 Gy (Scott et al., 1990), whereas the TD 50 for animals similarly exposed is 3.3 Gy (Diehl et al., 1992). (In this case, early death is used as the measure of toxicity for the purpose of determining the MTD.)
OCR for page 178
Issues in Risk Assessment
TABLE G-2 Deviations from Krewski's Regressions
Gold's Estimate (one-stage, time-corrected)
Krewski's Estimate (one-stage, time-corrected)
Krewski's Case No.
Chemical
MTD
log MTD
TD50
log TD50
Predicted log TD50
Difference
Standardized Difference
TD50
log TD50
Predicted log TD50
Difference
Standardized Difference
2
2-Acetyl-amino-fluorene
10.0
1.00
3.78
0.577
0.835
-0.318
-0.582
3.83
0.583
0.895
-0.312
-0.572
3
Acrylonitrile
5.69
0.755
5.31
0.725
0.647
0.078
0.144
5.61
0.755
0.647
0.102
0.187
1
Benzidine
80.0
1.903
8.99
0.954
1.811
-0.857
-1.571
9.10
0.959
1.811
-0.851
-1.561
3
C.I. Direct Black 38
60.0
1.778
0.945
-0.025
1.684
-1.708
-3.132
0.99
-0.002
1.684
-1.688
-3.095
6
C.I. Direct Blue 6
60.0
1.778
1.18
0.072
1.684
-1.612
-2.955
1.17
0.068
1.684
-1.616
-3.962
8
C.I. Direct Brown 95
75.0
1.875
2.07
0.316
1.782
-1.466
-2.688
2.62
0.418
1.782
-1.364
-2.500
3
CCl4
1650
3.217
114
2.057
3.143
-1.086
-1.991
115.12
2.061
3.143
-1.082
-1.983
3
EDB
28.1
1.449
1.26
0.100
1.350
-1.250
-2.291
5.60
0.748
1.350
-0.608
-1.103
1
ENU
0.429
-3.68
0.904
-0.044
-0.491
0.447
0.820
0.94
-0.026
-0.491
0.464
0.851
2
Ethylene
6.11
0.786
7.43
0.871
0.678
0.193
0.354
7.69
0.786
0.678
0.208
0.381
9
MOCA
34.0
1.531
20.8
1.318
1.434
-0.116
-0.212
21.07
1.324
1.434
-0.110
-0.202
13
Vinyl chloride (CPDB)
0.279
-0.554
14.2
1.152
-0.681
1.833
3.360
33.52
1.525
-0.681
2.206
4.044
OCR for page 179
Issues in Risk Assessment
Gold's Estimate (one-stage, time-corrected)
Krewski's Estimate (one-stage, time-corrected)
Krewski's Case No.
Chemical
MTD
log MTD
TD50
log TD50
Predicted log TD50
Difference
Standardized Difference
TD50
log TD50
Predicted log TD50
Difference
Standardized Difference
Chemicals or studies not in the CPDB or in Krewski's analysis
Benzo[a]-pyrene
32.5
1.512
11.70
1.068
1.414
-0.346
-0.634
1,3-Butadiene ()
380
2.580
20.81
1.318
2.496
-1.178
-2.160
Vinly Chloride (CRAM)
14.1
1.149
4.89
0.689
1.046
-0.357
-0.654
OCR for page 180
Issues in Risk Assessment
TABLE G-3 1,3 Butadiene*
Dose Rate (mg/kg-d)
Tumor Incidence
Males
Females
Lymphocytic lymphoma
0
2/70
2/70
3.8
1/70
4/70
12
2/70
6/70
38
4/70
3/70
120
2/70
11/70
380
62/90
36/90
*Inhalation exposure, 6h/day, 5d/wk for up to 2 years. Most animals died in high exposure groups by 65 weeks because of high tumor incidence.
Source: Melnick et al., 1990.
TABLE G-4 Benzo[a]pyrene*
Dose Rate
(mg/kg-d)
Tumor Incidence
Male and Female
Stomach, squamous cell carcinomas and papillomas
0
0/289
0.13
0/25
1.3
0/24
2.6
1/23
3.9
0/37
5.2
1/40
5.85
4/40
6.5
24/34
13.0
19/23
32.5
66/73
*Oral exposure in diet. Mice, CFW, male and female. Duration of exposure: 110 days. Duration of experiment: 183 days.
Source: Neal and Rigdon, 1967.
OCR for page 181
Issues in Risk Assessment
TABLE G-5 Vinyl Chloride*
Dose Rate (mg/kg-d)
Tumor Incidence
Males
Females
Liver tumors (neoplastic nodules, hepatocellular carcinomas, angiosarcomas)
0
0/55
2/57
1.7
2/58
26/58
5.0
17/56
42/59
14.1
58/59
56/57
*Oral lifetime exposure. Surviving males and females were sacrificed at 135 and 144 weeks, respectively.
Source: Feron et al., 1981.
RESULTS
For the 14 candidate agents on which comparable quantitative data are available, the right side of Table G-2 shows the observed TD50 as calculated by Krewski's procedures with the one-stage model. The last three columns in Table G-2 show the log10TD50 predicted by Krewski's model, the deviation from the regression line (observed - predicted log10TD50), and the standardized deviation (observed deviation divided by r.m.s. error). The data are plotted in relation to Krewski's regression line in Figure 1. Most of the 14 candidate chemicals are within the 95% confidence limits (standardized deviation, 1.96); this is illustrated in Figure 2, which plots the calculated TD50s for each of the 14 chemicals. For five of the 14 agents (the three benzidine dyes, carbon tetrachloride, and 1,3-butadiene), the calculated TD50s are below the lower confidence limit on the regression line, i.e., inside Region C (Figure 2). Among the four agents on which comparable quantitative data are not available, only plutonium has a low ratio of TD50 to HDT (1:13), but its HDT caused premature deaths and would not be accepted as an MTD in a conventional bioassay.
Neither of the TD50s calculated for vinyl chloride fall in Region C. The TD50 based on the data selected by the committee falls below the
OCR for page 182
Issues in Risk Assessment
regression line in Region B; the TD50 based on the data in the CPDB falls above the upper confidence limit on the regression line in Region A.
The TD50s in the CPDB for the 12 potential supercarcinogens already included in Krewski's 191 chemicals are generally very close to those calculated by Krewski. There are, however, two notable differences: EDB and vinyl chloride. The TD50 for EDB in the CPDB is based on lifetable methods, and is thus somewhat different than that calculated by Krewski using summary tumor incidence data. The discrepency between the Gold and Krewski TD50s for vinyl chloride based on the data from the CPDB is apparently due to differences in the numerical procedures used in model fitting. (This difference is small in relation to the wide variation in TD50s in the CPDB based on different experiments with vinyl chloride.) Neither of these differences is particularly relevant to the search for supercarcinogens because the TD50s for these two compounds do not fall in Region C.
DISCUSSION
The results just discussed do not provide strong evidence of the existence of supercarcinogens. Of the 14 chemicals considered as potential supercarcinogens, only five fall inside Region C; even these five are only slightly beyond the boundary separating regions B and C.
These results are based on certain assumptions about the appropriate adjustments to be applied to dose (and hence to potency) in experiments that are terminated substantially earlier than the 2 year lifetime of rodents. Those assumptions are based on sparse empirical evidence and are somewhat arbitrary. A common generalization is that cancer incidence is proportional to fn where the exponent n may range from 2 to 6 (Armitage and Doll, 1961). The CPDB's procedures are equivalent to the assumption that n = 2, which gives relatively low estimates of carcinogenic potency. An assumption that n = 3 or higher would shift the estimates of TD50 for the chemicals under review still further into Region C.
In summary, the results of the committee's informal study suggest that supercarcinogens are rare. The best candidates for designation as supercarcinogens are a few agents that induce cancer in rodents unusually
OCR for page 183
Issues in Risk Assessment
early in life. For such agents, the definition of potency is somewhat arbitrary: the more account that is taken of their early action, the higher the estimates of potency and the weaker the general relationship between potency and toxicity.
OCR for page 184
Issues in Risk Assessment
This page in the original is blank.
Representative terms from entire chapter:
tumor incidence
